High Hopes: Saving for High School with a Mobile Money Lock Box

Last registered on July 26, 2014

Pre-Trial

Trial Information

General Information

Title
High Hopes: Saving for High School with a Mobile Money Lock Box
RCT ID
AEARCTR-0000435
Initial registration date
July 26, 2014

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
July 26, 2014, 4:53 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Region

Primary Investigator

Affiliation
Georgetown University

Other Primary Investigator(s)

PI Affiliation
Georgetown University

Additional Trial Information

Status
In development
Start date
2014-07-14
End date
2015-12-31
Secondary IDs
Abstract
The transition from primary to high school can be costly, even when tuition fees are low or zero. In this project, undertaken in Kenya, we provide parents of soon-to-graduate primary school children with a variety of vehicles to save for these expenses, and measure their responses in terms of take-up, amounts saved, and enrollment and attendance at high school. The primary focus is on a mobile money lock-box, held within a bank account, itself linked to and accessed through a mobile money platform. Balances held to maturity in the lock box earn a bonus interest rate, but this bonus is lost on early withdrawals, which can only be accessed after a waiting period of 48 hours. We measure the net effect of the time penalty (the lock) and the interest rate bonus, by comparing a treatment group with a pure control group and with a group that is encouraged to save on their mobile phone, but not in the lock box. We cross the different mobile money product arms of our RCT with a text messaging intervention reminding users of the goals they set at recruitment.
External Link(s)

Registration Citation

Citation
Habyarimana, James and William Jack. 2014. "High Hopes: Saving for High School with a Mobile Money Lock Box." AEA RCT Registry. July 26. https://doi.org/10.1257/rct.435-1.0
Former Citation
Habyarimana, James and William Jack. 2014. "High Hopes: Saving for High School with a Mobile Money Lock Box." AEA RCT Registry. July 26. https://www.socialscienceregistry.org/trials/435/history/2231
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
This is an encouragement design in which we have two main treatment groups and a control. Parents of Standard 8 students in Kenyan primary schools attend meetings at which they are engaged in a discussion about saving for the transition to secondary school, which will occur in January 2015. All groups are given this information, and all parents are asked to estimate how much they would need to save in order to cover the costs of sending their child to high school. Our interventions include encouragement to use mobile phone based savings products, and exposure to text messages reminding parents of their savings goals.
Intervention Start Date
2014-07-14
Intervention End Date
2015-01-31

Primary Outcomes

Primary Outcomes (end points)
Key outcomes are financial balances on mobile phone accounts, and other savings levels and accumulation of assets. In addition, we will document whether the children of study participants enter high school, which school they attend, and their test scores at the end of primary school.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
We have two primary treatment groups and a control. The first main treatment group is encouraged to open a mobile phone based "lock savings account" from which balances cannot be withdrawn without penalty until January 2015. A second treatment group is encouraged to open a mobile phone based savings account that earns less interest but does not have any lock features. The control group receives no information about saving on their mobile phones.

In addition, we cross each group with a text message intervention that reminds parents of their specific savings goal, and how long they have left to achieve it.
Experimental Design Details
Randomization Method
Parents are interviewed in meetings held at 360 primary schools in three counties in Kenya. Randomization is at the school level stratified by county, using an office computer.
Randomization Unit
Randomization is at the school level.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
360 primary schools.
Sample size: planned number of observations
Parents of 9,000 students
Sample size (or number of clusters) by treatment arms
120 control, 120 mobile savings account, 120 mobile lock savings account
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We assume a two level cluster randomized trial 2-CRT with level 1 being the student and level 2 being the class/teacher and level 3 being the school. The key parameters are the share of between school variance (rho_S). We have no prior data on which to draw this estimate so we assume conservative range of [0.05 0.2]. We assume that net recruitment at the school produces 20 parents who sign up for the savings products. We also assume that baseline data can explain about 30% of the variance of the outcome (we test for sensitivity to this assumption by showing the MDE-sample size locus when there are no covariates). Given the assumptions above and a total sample size of 120 schools per arm, we have 80% power to distinguish effect sizes as small as 0.105 (at the low end of the ICC) or 0.156 at the high end of the ICC.   Two possible scenarios: Firstly that uptake or compliance is considerably lower than the 20 student/families. We revise our cluster size assumptions to 10 students per cluster maintaining the baseline ICC and baseline covariate explanatory power assumptions. Our MDE now has the following range [0.132 0.171]. Secondly, if we assume an even more conservative scenario; no baseline covariates and an ICC of 0.4, then our MDE increases to 0.24.
IRB

Institutional Review Boards (IRBs)

IRB Name
Innovations for Poverty Action
IRB Approval Date
2014-06-16
IRB Approval Number
3338
IRB Name
Georgetown University IRB
IRB Approval Date
2014-06-03
IRB Approval Number
2014-0222

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials